import pandas as pd

# Load the CSV file with semicolon as delimiter
data = pd.read_csv("Customer_Behavior_Data.csv", delimiter=";")

# Display the first few rows to check the structure
print(data.head())

   account length  location code  user id credit card info save push status  \
0             128            415  3824657                    no         yes   
1             107            415  3717191                    no         yes   
2             137            415  3581921                    no          no   
3              84            408  3759999                   yes          no   
4              75            415  3306626                   yes          no   

   add to wishlist  desktop sessions  app sessions  desktop transactions  \
0               25               265            45                    17   
1               26               162            27                    17   
2                0               243            41                    10   
3                0               299            51                     5   
4                0               167            28                    13   

   total product detail views  session duration  promotion clicks  \
0                         110               197                87   
1                         123               196               103   
2                         114               121               110   
3                          71                62                88   
4                         113               148               122   

  avg order value  sale product views discount rate per visited products  \
0           244,7                  91                              11,01   
1           254,4                 103                              11,45   
2           162,6                 104                               7,32   
3           196,9                  89                               8,86   
4           186,9                 121                               8,41   

  product detail view per app session  app transactions  \
0                                  10                 3   
1                                13,7                 3   
2                                12,2                 5   
3                                 6,6                 7   
4                                10,1                 3   

  add to cart per session  customer service calls  churn  
0                     2,7                       1      0  
1                     3,7                       1      0  
2                    3,29                       0      0  
3                    1,78                       2      0  
4                    2,73                       3      0

from google.colab import drive
drive.mount('/content/drive')

Mounted at /content/drive

data

# Get data types and check for missing values
data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 3333 entries, 0 to 3332
Data columns (total 20 columns):
 #   Column                               Non-Null Count  Dtype 
---  ------                               --------------  ----- 
 0   account length                       3333 non-null   int64 
 1   location code                        3333 non-null   int64 
 2   user id                              3333 non-null   int64 
 3   credit card info save                3333 non-null   object
 4   push status                          3333 non-null   object
 5   add to wishlist                      3333 non-null   int64 
 6   desktop sessions                     3333 non-null   int64 
 7   app sessions                         3333 non-null   int64 
 8   desktop transactions                 3333 non-null   int64 
 9   total product detail views           3333 non-null   int64 
 10  session duration                     3333 non-null   int64 
 11  promotion clicks                     3333 non-null   int64 
 12  avg order value                      3333 non-null   object
 13  sale product views                   3333 non-null   int64 
 14  discount rate per visited products   3333 non-null   object
 15  product detail view per app session  3333 non-null   object
 16  app transactions                     3333 non-null   int64 
 17  add to cart per session              3333 non-null   object
 18  customer service calls               3333 non-null   int64 
 19  churn                                3333 non-null   int64 
dtypes: int64(14), object(6)
memory usage: 520.9+ KB

# Check for missing values
data.isnull().sum()

# Basic statistics
data.describe().T

# Shape of the data
print("Data Shape:", data.shape)

# Unique values per column
unique_values = data.nunique()
unique_values

Data Shape: (3333, 20)

import matplotlib.pyplot as plt
import seaborn as sns

plt.figure(figsize=(8, 6))
sns.histplot(data['account length'], kde=True)
plt.title('Distribution of Account Length')
plt.xlabel('Account Length (days)')
plt.show()

plt.figure(figsize=(8, 6))
sns.histplot(data['desktop sessions'], color='blue', label='Desktop Sessions', kde=True)
sns.histplot(data['app sessions'], color='orange', label='App Sessions', kde=True)
plt.title('Desktop vs. App Sessions')
plt.xlabel('Number of Sessions')
plt.legend()
plt.show()

plt.figure(figsize=(8, 6))
sns.histplot(data['avg order value'], kde=True, color='green')
plt.title('Distribution of Average Order Value')
plt.xlabel('Average Order Value')
plt.show()

# Count plot for churn status
plt.figure(figsize=(8, 6))
sns.countplot(x='churn', data=data, palette='viridis')
plt.title('Churn Status')
plt.xlabel('Churn')
plt.ylabel('Number of Customers')
plt.show()

<ipython-input-13-1e512c34adce>:3: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.countplot(x='churn', data=data, palette='viridis')

# Count plots for each categorical variable
import matplotlib.pyplot as plt
import seaborn as sns

categorical_cols = ['credit card info save', 'push status']  # Replace with actual categorical column names

for col in categorical_cols:
    plt.figure(figsize=(8, 5))
    sns.countplot(data=data, x=col, palette="viridis")
    plt.title(f"Count of {col}")
    plt.show()

<ipython-input-14-c5a6e8b1d039>:9: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.countplot(data=data, x=col, palette="viridis")

<ipython-input-14-c5a6e8b1d039>:9: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.countplot(data=data, x=col, palette="viridis")

# Pair plot for key numerical relationships, separated by churn status
sns.pairplot(data, hue='churn', palette="husl", diag_kind="kde")  # 'churn' is the target column
plt.suptitle("Pairwise Plot by Churn Status", y=1.02)
plt.show()

# Violin plot to compare distributions of app and desktop metrics by churn status
app_metrics = ['app sessions', 'app transactions']
desktop_metrics = ['desktop sessions', 'desktop transactions']

for col in app_metrics + desktop_metrics:
    plt.figure(figsize=(8, 5))
    sns.violinplot(data=data, x='churn', y=col, palette="muted")
    plt.title(f"{col} Distribution by Churn Status")
    plt.show()

<ipython-input-16-00ccb04ca475>:7: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.violinplot(data=data, x='churn', y=col, palette="muted")

<ipython-input-16-00ccb04ca475>:7: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.violinplot(data=data, x='churn', y=col, palette="muted")

<ipython-input-16-00ccb04ca475>:7: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.violinplot(data=data, x='churn', y=col, palette="muted")

<ipython-input-16-00ccb04ca475>:7: FutureWarning: 

Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.

  sns.violinplot(data=data, x='churn', y=col, palette="muted")

import numpy as np
# Plotting the correlation matrix
plt.figure(figsize=(12, 10))
sns.heatmap(data.select_dtypes(include=[np.number]).corr(), annot=True, cmap="coolwarm", fmt=".3f")
plt.title("Correlation Matrix of Numerical Columns")
plt.show()

summary_stats = data.describe().T
summary_stats['skew'] = data.select_dtypes(include=[np.number]).skew()  # Skewness shows asymmetry of distribution

summary_stats

from sklearn.preprocessing import LabelEncoder

# Initialize the LabelEncoder
label_encoder = LabelEncoder()

# Apply LabelEncoder on 'credit card info save' and 'push status'
data['credit card info save'] = label_encoder.fit_transform(data['credit card info save'])
data['push status'] = label_encoder.fit_transform(data['push status'])

# Display the first few rows to confirm the encoding
data.head()

# Box plot for key activity metrics
columns_to_plot = ['desktop sessions', 'app sessions', 'desktop transactions',
                   'app transactions', 'session duration', 'promotion clicks']
for column in columns_to_plot:
    plt.figure(figsize=(8, 4))
    sns.boxplot(data[column])
    plt.title(f'Box Plot of {column}')
    plt.show()

# Scatter plot of session duration vs. desktop transactions
plt.figure(figsize=(10, 6))
sns.scatterplot(data=data, x='session duration', y='desktop transactions', hue='churn')
plt.title('Session Duration vs. Desktop Transactions')
plt.xlabel('Session Duration')
plt.ylabel('Desktop Transactions')
plt.show()

# Pair plot of selected features
sns.pairplot(data[['desktop sessions', 'app sessions', 'promotion clicks', 'session duration', 'churn']], hue='churn')
plt.suptitle('Pair Plot of Selected Features (Desktop, App Sessions, Promotions, Churn)', y=1.02)
plt.show()

# Bar plot of average transactions by churn status
plt.figure(figsize=(8, 5))
sns.barplot(data=data, x='churn', y='desktop transactions', estimator='mean', ci=None)
plt.title('Average Desktop Transactions by Churn Status')
plt.xlabel('Churn')
plt.ylabel('Average Desktop Transactions')
plt.show()

<ipython-input-23-016eed8ca896>:3: FutureWarning: 

The `ci` parameter is deprecated. Use `errorbar=None` for the same effect.

  sns.barplot(data=data, x='churn', y='desktop transactions', estimator='mean', ci=None)

# Scatter plot for session duration vs. desktop and app transactions
plt.figure(figsize=(12, 6))
sns.scatterplot(data=data, x='session duration', y='desktop transactions', hue='churn')
plt.title('Relationship between Session Duration and Desktop Transactions')
plt.xlabel('Session Duration (minutes)')
plt.ylabel('Desktop Transactions')
plt.show()

plt.figure(figsize=(12, 6))
sns.scatterplot(data=data, x='session duration', y='app transactions', hue='churn')
plt.title('Relationship between Session Duration and App Transactions')
plt.xlabel('Session Duration (minutes)')
plt.ylabel('App Transactions')
plt.show()

# Scatter plot for wishlist additions vs. transactions
plt.figure(figsize=(10, 6))
sns.scatterplot(data=data, x='add to wishlist', y='desktop transactions', hue='churn')
plt.title('Wishlist Additions vs. Desktop Transactions')
plt.xlabel('Add to Wishlist Count')
plt.ylabel('Desktop Transactions')
plt.show()

plt.figure(figsize=(10, 6))
sns.scatterplot(data=data, x='add to wishlist', y='app transactions', hue='churn')
plt.title('Wishlist Additions vs. App Transactions')
plt.xlabel('Add to Wishlist Count')
plt.ylabel('App Transactions')
plt.show()

# Scatter plot for promotion clicks vs. desktop and app transactions
plt.figure(figsize=(10, 6))
sns.scatterplot(data=data, x='promotion clicks', y='desktop transactions', hue='churn')
plt.title('Promotion Clicks vs. Desktop Transactions')
plt.xlabel('Promotion Clicks')
plt.ylabel('Desktop Transactions')
plt.show()

plt.figure(figsize=(10, 6))
sns.scatterplot(data=data, x='promotion clicks', y='app transactions', hue='churn')
plt.title('Promotion Clicks vs. App Transactions')
plt.xlabel('Promotion Clicks')
plt.ylabel('App Transactions')
plt.show()

# Box plot of customer service calls by churn status
plt.figure(figsize=(8, 5))
sns.boxplot(data=data, x='churn', y='customer service calls')
plt.title('Customer Service Calls by Churn Status')
plt.xlabel('Churn Status')
plt.ylabel('Customer Service Calls')
plt.show()

data

# Convert commas to periods
converted_data = data["discount rate per visited products"].replace(',', '.')

# Print the result
print(converted_data)

0       11,01
1       11,45
2        7,32
3        8,86
4        8,41
        ...  
3328    12,56
3329     8,61
3330     8,64
3331     6,26
3332    10,86
Name: discount rate per visited products, Length: 3333, dtype: object

data

data.to_csv('data_new.csv', index=False)

# Remove all commas from the data
df = data.applymap(lambda x: str(x).replace(',', '') if isinstance(x, str) else x)

<ipython-input-35-5210660e964e>:2: FutureWarning: DataFrame.applymap has been deprecated. Use DataFrame.map instead.
  df = data.applymap(lambda x: str(x).replace(',', '') if isinstance(x, str) else x)

df

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 3333 entries, 0 to 3332
Data columns (total 20 columns):
 #   Column                               Non-Null Count  Dtype 
---  ------                               --------------  ----- 
 0   account length                       3333 non-null   int64 
 1   location code                        3333 non-null   int64 
 2   user id                              3333 non-null   int64 
 3   credit card info save                3333 non-null   int64 
 4   push status                          3333 non-null   int64 
 5   add to wishlist                      3333 non-null   int64 
 6   desktop sessions                     3333 non-null   int64 
 7   app sessions                         3333 non-null   int64 
 8   desktop transactions                 3333 non-null   int64 
 9   total product detail views           3333 non-null   int64 
 10  session duration                     3333 non-null   int64 
 11  promotion clicks                     3333 non-null   int64 
 12  avg order value                      3333 non-null   object
 13  sale product views                   3333 non-null   int64 
 14  discount rate per visited products   3333 non-null   object
 15  product detail view per app session  3333 non-null   object
 16  app transactions                     3333 non-null   int64 
 17  add to cart per session              3333 non-null   object
 18  customer service calls               3333 non-null   int64 
 19  churn                                3333 non-null   int64 
dtypes: int64(16), object(4)
memory usage: 520.9+ KB

# Convert columns with numeric strings to integers
for col in df.columns:
    if df[col].dtype == 'object':
        df[col] = pd.to_numeric(df[col], errors='coerce')

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 3333 entries, 0 to 3332
Data columns (total 20 columns):
 #   Column                               Non-Null Count  Dtype
---  ------                               --------------  -----
 0   account length                       3333 non-null   int64
 1   location code                        3333 non-null   int64
 2   user id                              3333 non-null   int64
 3   credit card info save                3333 non-null   int64
 4   push status                          3333 non-null   int64
 5   add to wishlist                      3333 non-null   int64
 6   desktop sessions                     3333 non-null   int64
 7   app sessions                         3333 non-null   int64
 8   desktop transactions                 3333 non-null   int64
 9   total product detail views           3333 non-null   int64
 10  session duration                     3333 non-null   int64
 11  promotion clicks                     3333 non-null   int64
 12  avg order value                      3333 non-null   int64
 13  sale product views                   3333 non-null   int64
 14  discount rate per visited products   3333 non-null   int64
 15  product detail view per app session  3333 non-null   int64
 16  app transactions                     3333 non-null   int64
 17  add to cart per session              3333 non-null   int64
 18  customer service calls               3333 non-null   int64
 19  churn                                3333 non-null   int64
dtypes: int64(20)
memory usage: 520.9 KB

# Define features and target variable
X = df.drop(columns=['churn', 'user id'])  # Drop target and ID columns
y = df['churn']

# Split data into training and test sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)

# Initialize and train the Random Forest model
model = RandomForestClassifier(n_estimators=100, random_state=42)
model.fit(X_train, y_train)

RandomForestClassifier(random_state=42)

RandomForestClassifier(random_state=42)

# Predict and evaluate the model
y_pred = model.predict(X_test)
print(classification_report(y_test, y_pred))

              precision    recall  f1-score   support

           0       0.95      1.00      0.97       857
           1       0.96      0.68      0.80       143

    accuracy                           0.95      1000
   macro avg       0.95      0.84      0.88      1000
weighted avg       0.95      0.95      0.95      1000

	count	mean	std	min	25%	50%	75%	max
account length	3333.0	1.010648e+02	39.822106	1.0	74.0	101.0	127.0	243.0
location code	3333.0	4.371824e+02	42.371290	408.0	408.0	415.0	510.0	510.0
user id	3333.0	3.746291e+06	274662.573752	3271058.0	3508680.0	3748187.0	3985970.0	4229964.0
add to wishlist	3333.0	8.099010e+00	13.688365	0.0	0.0	0.0	20.0	51.0
desktop sessions	3333.0	1.798119e+02	54.457135	0.0	144.0	179.0	216.0	351.0
app sessions	3333.0	3.056796e+01	9.269376	0.0	24.0	31.0	37.0	60.0
desktop transactions	3333.0	1.708761e+01	4.323795	0.0	14.0	17.0	20.0	31.0
total product detail views	3333.0	1.004356e+02	20.069084	0.0	87.0	101.0	114.0	165.0
session duration	3333.0	2.010396e+02	50.714359	0.0	167.0	201.0	235.0	364.0
promotion clicks	3333.0	1.001107e+02	19.923911	0.0	87.0	100.0	114.0	170.0
sale product views	3333.0	1.001077e+02	19.568609	33.0	87.0	100.0	113.0	175.0
app transactions	3333.0	4.479448e+00	2.461214	0.0	3.0	4.0	6.0	20.0
customer service calls	3333.0	1.562856e+00	1.315491	0.0	1.0	1.0	2.0	9.0
churn	3333.0	1.449145e-01	0.352067	0.0	0.0	0.0	0.0	1.0

	count	mean	std	min	25%	50%	75%	max	skew
account length	3333.0	1.010648e+02	39.822106	1.0	74.0	101.0	127.0	243.0	0.096606
location code	3333.0	4.371824e+02	42.371290	408.0	408.0	415.0	510.0	510.0	1.126823
user id	3333.0	3.746291e+06	274662.573752	3271058.0	3508680.0	3748187.0	3985970.0	4229964.0	0.009732
add to wishlist	3333.0	8.099010e+00	13.688365	0.0	0.0	0.0	20.0	51.0	1.264824
desktop sessions	3333.0	1.798119e+02	54.457135	0.0	144.0	179.0	216.0	351.0	-0.028737
app sessions	3333.0	3.056796e+01	9.269376	0.0	24.0	31.0	37.0	60.0	-0.028420
desktop transactions	3333.0	1.708761e+01	4.323795	0.0	14.0	17.0	20.0	31.0	-0.010819
total product detail views	3333.0	1.004356e+02	20.069084	0.0	87.0	101.0	114.0	165.0	-0.111787
session duration	3333.0	2.010396e+02	50.714359	0.0	167.0	201.0	235.0	364.0	-0.024248
promotion clicks	3333.0	1.001107e+02	19.923911	0.0	87.0	100.0	114.0	170.0	-0.055096
sale product views	3333.0	1.001077e+02	19.568609	33.0	87.0	100.0	113.0	175.0	0.032500
app transactions	3333.0	4.479448e+00	2.461214	0.0	3.0	4.0	6.0	20.0	1.321478
customer service calls	3333.0	1.562856e+00	1.315491	0.0	1.0	1.0	2.0	9.0	1.091359
churn	3333.0	1.449145e-01	0.352067	0.0	0.0	0.0	0.0	1.0	2.018356

	account length	location code	user id	credit card info save	push status	add to wishlist	desktop sessions	app sessions	desktop transactions	total product detail views	session duration	promotion clicks	avg order value	sale product views	discount rate per visited products	product detail view per app session	app transactions	add to cart per session	customer service calls	churn
0	128	415	3824657	no	yes	25	265	45	17	110	197	87	244,7	91	11,01	10	3	2,7	1	0
1	107	415	3717191	no	yes	26	162	27	17	123	196	103	254,4	103	11,45	13,7	3	3,7	1	0
2	137	415	3581921	no	no	0	243	41	10	114	121	110	162,6	104	7,32	12,2	5	3,29	0	0
3	84	408	3759999	yes	no	0	299	51	5	71	62	88	196,9	89	8,86	6,6	7	1,78	2	0
4	75	415	3306626	yes	no	0	167	28	13	113	148	122	186,9	121	8,41	10,1	3	2,73	3	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
3328	192	415	4144276	no	yes	36	156	27	18	77	216	126	279,1	83	12,56	9,9	6	2,67	2	0
3329	68	415	3703271	no	no	0	231	39	13	57	153	55	191,3	123	8,61	9,6	4	2,59	3	0
3330	28	510	3288230	no	no	0	181	31	25	109	289	58	191,9	91	8,64	14,1	6	3,81	2	0
3331	184	510	3646381	yes	no	0	214	36	14	105	160	84	139,2	137	6,26	5	10	1,35	2	0
3332	74	415	4004344	no	yes	25	234	40	23	113	266	82	241,4	77	10,86	13,7	4	3,7	0	0

	account length	location code	user id	credit card info save	push status	add to wishlist	desktop sessions	app sessions	desktop transactions	total product detail views	session duration	promotion clicks	avg order value	sale product views	discount rate per visited products	product detail view per app session	app transactions	add to cart per session	customer service calls	churn
0	128	415	3824657	0	1	25	265	45	17	110	197	87	2447	91	1101	10	3	27	1	0
1	107	415	3717191	0	1	26	162	27	17	123	196	103	2544	103	1145	137	3	37	1	0
2	137	415	3581921	0	0	0	243	41	10	114	121	110	1626	104	732	122	5	329	0	0
3	84	408	3759999	1	0	0	299	51	5	71	62	88	1969	89	886	66	7	178	2	0
4	75	415	3306626	1	0	0	167	28	13	113	148	122	1869	121	841	101	3	273	3	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
3328	192	415	4144276	0	1	36	156	27	18	77	216	126	2791	83	1256	99	6	267	2	0
3329	68	415	3703271	0	0	0	231	39	13	57	153	55	1913	123	861	96	4	259	3	0
3330	28	510	3288230	0	0	0	181	31	25	109	289	58	1919	91	864	141	6	381	2	0
3331	184	510	3646381	1	0	0	214	36	14	105	160	84	1392	137	626	5	10	135	2	0
3332	74	415	4004344	0	1	25	234	40	23	113	266	82	2414	77	1086	137	4	37	0	0